• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Best Horn Driver for Compressed Recordings?

jkff

Active Member
I'm not using any Klipsch enclosure designs, horns or drivers for what I'm looking to build. But unlike the Lansing section of this forum, many here have had great success with the B&C DCM50.

However, as discussed here, from what Troy Crowe measured and what he and those here have heard from it, might the DCM50 be the best driver for my compressed recordings? https://audiokarma.org/forums/index...orn-driver-for-compressed-recordings.1058035/

Or might there be an even better driver for same?
 
Register to hide this ad
I'm not using any Klipsch enclosure designs, horns or drivers for what I'm looking to build. But unlike the Lansing section of this forum, many here have had great success with the B&C DCM50.

However, as discussed here, from what Troy Crowe measured and what he and those here have heard from it, might the DCM50 be the best driver for my compressed recordings? https://audiokarma.org/forums/index...orn-driver-for-compressed-recordings.1058035/

Or might there be an even better driver for same?
Yes.
 
Combination of horn and driver together, and how well they integrate together with the rest of the system - will determine the outcome.

A phenolic or cloth diaphragm can be more forgiving, but not on a bad horn or with a poorly designed network.

Garbage in garbage out.
 
Last edited:
Combination of horn and driver together, and how well they are integrated together and the rest of the system will determine the outcome.

A phenolic or cloth diaphragm can be more forgiving, but not on a bad horn or with a poorly designed network.

Garbage in garbage out.
For the DCM50 above my 416s
,

one of these horns?

or if the JMLC 340 horn would be a better choice, why?
 
Unless you find someone with the same driver / horn combo and get feedback from them, it is impossible to say what driver horn / combo will sound like. The crossover design will also make the driver / horn combo have a distinct sonic signature too.
 
Just realized you’re doing a two-way solution. Unless you are going active, you will need a tweeter as well. That DCM50 falls like a rock at 10kHz.
 
Just realized you’re doing a two-way solution. Unless you are going active, you will need a tweeter as well. That DCM50 falls like a rock at 10kHz.
Actually, because those speakers would be used primarily for playing many of my favorite recordings, but which were deliberately hit with excessive compression in mastering room to make them usable on the cheap vinyl players of the day, I may have my builder Troy Crowe omit a tweeter-unless upon playing excerpts of several of my uncompressed WAV files of CD track rips from my Google Drive, he would instead recommend including it. You ought to check this out from 17:15 to end.
 
...However, as discussed here, from what Troy Crowe measured and what he and those here have heard from it, might the DCM50 be the best driver for my compressed recordings?...
I know one thing here to be true: once recordings are compressed during the mixing and/or mastering process, there really is nothing that can be done to "undo" that compression.

You can use a multiband upward expander (usually called a multiband compressor/expander) to recover some of what was lost in the recordings themselves. But in my experience with this process has not been terribly successful on heavily compressed recordings.

To believe that by changing your loudspeaker's horns and/or compression drivers to accomplish the same effect as de-compressing your compressed music tracks--is like thinking that you can extract gold from garbage...which is "magical thinking"...at best.

The best thing that you can do for improved playback of compressed tracks is to flatten the transfer function (SPL and phase response) of your loudspeakers--that I've found--and to reduce in-room early acoustic reflections. I recommend measuring what you have using something like REW and applying inverse transfer function via DSP crossover (or upstream DSP filtering) out, particularly using FIR filtering to flatten both SPL and phase response.

I also recommend using demastering methods to correct some of the SPL imbalances put on the recordings themselves by mixing and mastering guys, and reduce any droning noise--particularly 50/60 Hz AC line noise and harmonics. I recommend Audacity to do that task.

Chris
 
Thanks for this. I glanced at your system on your profile page. I might want to drive active crossovers eventually, but I want to try the other suggested first. Have you ever applied those solutions to any of your favorite but compressed/limited dynamic range recordings? Percentage wise, how much better did it sound?
 
...Percentage wise, how much better did it sound?

I've found that the results of demastering my music tracks to be all over the map: some improved dramatically...and others, not as much. But I've found that over 90% of my recorded music library is noticeably improved using this technique--to the point that I generally demaster (in some way) virtually every newly acquired CD that comes into my library.

SACD disc tracks cannot be demastered (i.e., ripped DSD files), and DVD-As and Blu-Ray PCM music multichannel discs are sometimes okay as-is, or sometimes heavily clipped by the mastering people in order to make the tracks sound louder directly off the disc. Running Clip-Fix, then Normalize... on these generally reconstructs most of the clipped peaks and returns the overall track loudness back to as-recorded levels (i.e., near zero on the ReplayGain scale), and eliminates the harshness and induced odd-order harmonics due to the clipped peaks from the released mastering effects. Demastering processes use the inherent precision of numerical methods, not the inherent fuzziness of analog reconstruction processes.

I've applied these processes using Audacity freeware to many thousands of music tracks, mostly ripped from older used CDs--but also for a handful of vinyl rips, which are usually plagued by much noisier conditions, but sometimes have a much higher dynamic range remaining particularly on popular recordings made since 1991. Some older recordings are just not available on CD (digital format).

The best results always are gleaned from the best quality CD tracks, which are also usually the oldest CD versions with the highest retained dynamic range, and were generally released before 1991. These are also the least expensive to buy used on outlets like Amazon Marketplace and Discogs. These discs generally sell below $6 USD/disc nowadays, but as recently as four years ago, these prices were below $4/disc (i.e., the shipping & handling cost only). A little digging on Discogs and the DR-Database to find the highest retained dynamic range disc version available is the ticket to much improved listening pleasure.

On a time-spent basis per disc, I can now stay slightly ahead of the playback of the tracks ripped from CD while doing initial (coarse correction) demastering of each individual CD track using Audacity in real time. But note that I've had lots of experience demastering tracks--about 10 years and perhaps 25-30 thousand individual tracks completed, probably in excess of 2000 hours/year spent in demastering music tracks.

For popular music genres, generally each music track must be demastered individually, while classical music discs generally can be demastered using a single demastering EQ curve (after declipping, which is done first via macro--called a "chain" in Audacity). Jazz discs are somewhere in between classical and popular discs in terms of using single custom demastering EQ curves for the entire disc or by individually derived EQ tracks for each track.

But I strongly recommend getting the SPL and phase response of your loudspeakers under control first, since this has such a strong effect on the subjective listening results, particularly when using horn-loaded loudspeakers in getting rid of the all-pass phase growth of the loudspeaker's crossover filters and time alignment of each loudspeaker "way", This is basically to achieve minimum phase loudspeaker response first before demastering. Otherwise, you'll be finding yourself re-doing your demastered tracks again to rebalance the resulting bass response, which is >3 dB more subjectively apparent on minimum phase loudspeakers in my experience.

Chris
 
Once the mastering guy uses compression on the tracks, there really is no going backwards to recapture the lost dynamic range.

The best approach is in finding the recordings with the highest retained dynamic range (i.e., the oldest CDs you can find), then removing what I'd term the "ham-handed mastering practices" from that older era of mastering (pre-1991) that primarily used EQ attenuation of bass below ~100 Hz to make the tracks sound louder on-disc.

Modern recordings are so damaged by newer practices that the only course of action is to find vinyl records that have higher retained dynamic range--because the mastering guys cannot clip the music peaks on vinyl to the extent that they use on digital. The needle simply jumps out of the groove if they try. It's the limitation of the physical medium itself which prevents the mastering people from further destroying the music enjoyment.

The downside of vinyl, of course, is:

1) cost (nowadays for newly released material),
2) ticks and pops,
3) greatly increased noise floor and modulation distortion, and
4) the time it takes to successfully rip the tracks from the record surface.

I find it easier (generally speaking) to avoid music genres whose recordings are extremely compressed. This seems to be the most successful approach.

Chris
 
Last edited:

Just wanted to say, nice seeing you here! I do a drive by every now & then and stumbled onto this thread. Good to see you and hope things are going well.

(now I get to go back out to find a 3/8" x 4 1/2" bolt with (as I recall) 20, maybe 24 threads/inch..... so far, I'm 0-3 in hardware stores)
 
I find it easier (generally speaking) to avoid music genres whose recordings are extremely compressed. This seems to be the most successful approach.

Chris
I agree with all of your answers to what they, effectively, deem as the wrong question from the OP.
 
The best thing that you can do for improved playback of compressed tracks is to flatten the transfer function (SPL and phase response) of your loudspeakers--that I've found--and to reduce in-room early acoustic reflections. I recommend measuring what you have using something like REW and applying inverse transfer function via DSP crossover (or upstream DSP filtering) out, particularly using FIR filtering to flatten both SPL and phase response.

But I strongly recommend getting the SPL and phase response of your loudspeakers under control first, since this has such a strong effect on the subjective listening results, particularly when using horn-loaded loudspeakers in getting rid of the all-pass phase growth of the loudspeaker's crossover filters and time alignment of each loudspeaker "way", This is basically to achieve minimum phase loudspeaker response first before demastering. Otherwise, you'll be finding yourself re-doing your demastered tracks again to rebalance the resulting bass response, which is >3 dB more subjectively apparent on minimum phase loudspeakers in my experience.

I have a limited theoretical grasp on what electroacoustic problems you're describing here, which seem to be related to crossover design, the problems with room modes themselves and how they interact with a crossover network’s phase response , driver time alignment-and possibly phase problems of some number of stereo recordings, but which (presumably) can be easily fixed with tools like this. https://www.izotope.com/en/products/rx/features/phase.html

My builder Troy Crowe (above videos) is aware that I do hope to play these trouble recordings on whatever driver/horn combination he suggests to complete these speakers. I will certainly ask him to look at your suggestions here as they impact crossover design-and for making these compressed recordings however more listenable.

As for trying to "demaster" recordings to at least give the illusion of restored dynamics to these recordings, many at Steve Hoffman forums have also suggested that it's normally best to pursue the earliest CD edition (on a reputable label) of an album (somehow my 1989 MCA issue of Kenny Rogers & the First Edition Greatest Hits came quickest to mind). Before ever upgrading my vinyl playing hardware, I walked away from vinyl since the 90s. Much of my 60s music I have on among later 80s to mid 90s released CDs and/or from the original labels, such as https://www.discogs.com/release/16230909-Gary-Lewis-The-Playboys-The-Legendary-Masters-Series

And then some of perhaps questionable quality.
https://www.discogs.com/release/4882520-Peter-Gordon-The-Best-Of-Peter-Gordon

But among my worst upsets is the most sizable collection of its kind to date.
https://www.discogs.com/release/351...e-Outer-Limits-Original-Television-Soundtrack

I've little reason to suspect soundtrack label La La Land Records of pursuing anything but the best quality analog or digital masters. None of these recordings were ever issued on vinyl and nothing has been issued since this release. And as discussed below, the only prior CD release was somehow hit with very audible clipping distortion, most likely during digital mastering by one of those two engineers. Hard to believe that it might be Bernie Grundman, but who knows. https://uadforum.com/community/inde...g-and-or-mastering-ca-1960.71207/#post-524968

But the damage common to both releases-excessive compression-was likely done long ago in the mastering room.

However, beyond the above precautions my builder take can heed for designing my new speakers, assuming I can’t attempt your restoration techniques with my old version of Samplitude https://www.soundonsound.com/reviews/magix-samplitude-pro-x

I’ll switch to Audacity. https://www.pcmag.com/reviews/audacity I also have version 5 of this

https://www.izotope.com/en/shop/rx-10-advanced/
 
I have rethought the question as I remembered a Klipsch rep who replaced his mid driver and tweeter on his Kornerhorns. You want a driver with lightest most rigid diaphragms you can find to move the resonances out side the audio bands of your crossovers. So maybe beryllium. I know he bought some Altec adapters to mount the JBL small format driver to the the treaded Klipsch 400 HZ horn. As I remember, faintly he used 2405's to replace the T-35. But that was so long ago. Middle to late 70's. Atlas and the EV tweeter used heavy phenolic diaphragms which I never liked either by Klipsch, Atlas, JBL, EV, Community, or Altec. Aluminum all the way at that time.
 
My builder Troy Crowe (above videos) is aware that I do hope to play these trouble recordings on whatever driver/horn combination he suggests to complete these speakers. I will certainly ask him to look at your suggestions here as they impact crossover design-and for making these compressed recordings however more listenable.
We've dialogued in the past. I have some real reservations from a prior dialogue we had on another audio forum. I experienced some of his carelessness about some of his comparative measurements of drivers and horns, his choice of horn profiles (and in particular the rationale for his choices), some of his cause/effect reasoning, and of course, his final recommendations (to put it mildly).

I discussed this with him. When he failed to adequately clear up his forum comments--that he acknowledged were in error privately--I now judge his opinions and methods to be highly suspect, I'm sorry to say. I would also be extremely concerned if he were the one that led you to believe that changing out horns/drivers will fix the subjective listening issues of compressed recordings. (In fact, I would go a bit further than that--be that as it may.)

But he seems to have a better than average grasp of a few technical issues and measurement techniques/types. To say it succinctly: his opinions are a very mixed bag. I actually don't trust anything he's written. That especially includes his interpretations of his measurements.

I have a limited theoretical grasp on what electroacoustic problems you're describing here, which seem to be related to crossover design, the problems with room modes themselves and how they interact with a crossover network’s phase response , driver time alignment-and possibly phase problems of some number of stereo recordings, but which (presumably) can be easily fixed with tools like this. https://www.izotope.com/en/products/rx/features/phase.html

Passive crossover filters all have one issue--they delay the lower frequency drivers relative to the higher frequency drivers in time. Since this delay is a function of the median crossover frequency of the two "ways" (higher and lower acoustic drivers being crossed), the discussion always turns to the subject of phase. A first order crossover filter set (high pass and low pass filters) will produce 90 degrees of phase lag on the lower frequency drivers, a second order filter set will produce 180 degrees of delay, etc. The only way to avoid this is to not use passive crossover filters, instead to use digital (a.k.a., DSP) active crossovers that can remove this delay. Mounting the higher frequency drivers slightly farther away from the listener can somewhat compensate for this effect, as has been used by many loudspeaker manufacturers. These low-frequency delays turn out to affect the subjective listening preferences of loudspeakers. But this is not a subject that is widely known among DIY loudspeaker hobbyists, and even loudspeaker design engineers in general. I documented my experiences in the "Subconscious Auditory Effects of Quasi-Linear Phase Loudspeakers" thread that I linked (and you copied again back to me).

As for trying to "demaster" recordings to at least give the illusion of restored dynamics to these recordings, many at Steve Hoffman forums have also suggested that it's normally best to pursue the earliest CD edition (on a reputable label) of an album (somehow my 1989 MCA issue of Kenny Rogers & the First Edition Greatest Hits came quickest to mind). Before ever upgrading my vinyl playing hardware, I walked away from vinyl since the 90s. Much of my 60s music I have on among later 80s to mid 90s released CDs and/or from the original labels, such as https://www.discogs.com/release/16230909-Gary-Lewis-The-Playboys-The-Legendary-Masters-Series
Unfortunately, I don't own either of those albums. Here is a list of my CDs and other music discs as of a couple of years ago. (If you have trouble accessing that link, let me know.) Perhaps there is something there that you also own and listen to. I might be able to provide an example of demastered track(s) for your review.

Of the Kenny Rogers album, look at the two different versions found on the DRDatabase site (bigger numbers are better): https://dr.loudness-war.info/?artist=Kenny+Rogers&album=Greatest

The version from 1988 is audibly more dynamic than the 1996 version. I recommend looking around that database site for other recordings and their corresponding versions/dynamic range ratings.

Chris
 
I have rethought the question as I remembered a Klipsch rep who replaced his mid driver and tweeter on his Kornerhorns. You want a driver with lightest most rigid diaphragms you can find to move the resonances out side the audio bands of your crossovers. So maybe beryllium. I know he bought some Altec adapters to mount the JBL small format driver to the the treaded Klipsch 400 HZ horn. As I remember, faintly he used 2405's to replace the T-35. But that was so long ago. Middle to late 70's. Atlas and the EV tweeter used heavy phenolic diaphragms which I never liked either by Klipsch, Atlas, JBL, EV, Community, or Altec. Aluminum all the way at that time.
I doubt any flaws in a recording will go unexposed by most if not all beryllium drivers, such as those Pierre used. https://www.usspeaker.com/radian 745neoBepb-1.htm
See posts #15,266, 15,276. https://www.diyaudio.com/community/threads/beyond-the-ariel.100392/page-764
The best thing that you can do for improved playback of compressed tracks is to flatten the transfer function (SPL and phase response) of your loudspeakers--that I've found--and to reduce in-room early acoustic reflections. I recommend measuring what you have using something like REW and applying inverse transfer function via DSP crossover (or upstream DSP filtering) out, particularly using FIR filtering to flatten both SPL and phase response.
But I strongly recommend getting the SPL and phase response of your loudspeakers under control first, since this has such a strong effect on the subjective listening results, particularly when using horn-loaded loudspeakers in getting rid of the all-pass phase growth of the loudspeaker's crossover filters and time alignment of each loudspeaker "way", This is basically to achieve minimum phase loudspeaker response first before demastering. Otherwise, you'll be finding yourself re-doing your demastered tracks again to rebalance the resulting bass response, which is >3 dB more subjectively apparent on minimum phase loudspeakers in my experience.


Here’s Pierre’s crossover and/or voicing scheme. https://www.diyaudio.com/community/threads/beyond-the-ariel.100392/page-727 See posts # 29, 36, 39.

https://www.diyaudio.com/community/...e-medium-high-efficiency-system.405604/page-2

And Pierre perfected the throat adapters for the Radian 745neoBe/Athos TH4001; posts 228, 234 -240.

https://www.diyaudio.com/community/...orns-measurements.395046/page-12#post-7350302

Pierre had generously reviewed most of my uncompressed WAV files of CD track rips via my Google Drive (and playlist/review sheet Excel file). While the sound quality of 4 or 5 tracks blew him away, most he found moderately to badly compressed.

I had hoped that the warm sound and/or other aspects of the B&C DCM50, as Troy Crowe described in his listening tests (after adding a ~6 uF cap) in the above video, might accentuate the compressed sound of most of my recordings a lot less than Pierre’s beryllium drivers. Or even a pair of these.

In any case, NO, neither Troy nor Pierre even speculated, , much less suggested, that any particular driver or horn would make my compressed recordings sound better.

So, as you say, choosing between those or any drivers is evidently no solution. Instead, trying to access among the earliest CD releases of the first generation analog master of a vintage recording and then applying prescribed “mastering” techniques to compensate for the (however unintentionally) destructive effects of excessively applied compression and eq. https://community.klipsch.com/index...taves-audacity-remastering-to-restore-tracks/

But presumably before proceeding with source material restoration, you strongly recommend achieving a “flat” transfer function result between SPL and phase response. https://www.roomeqwizard.com/help/help_en-GB/html/primer.html

Meanwhile, I am struggling badly trying to get through this.
https://community.klipsch.com/index...ld-you-explain-minimum-phase-analysis-please/

https://techtalk.parts-express.com/forum/tech-talk-forum/1290936-explaining-minimum-phase/page2

And even though all of my music sources are digital and my pc will be central to the 2.3 or 3.3 HT system Troy will build for me
(https://ark.intel.com/content/www/u...-1390-processor-16m-cache-up-to-5-20-ghz.html + 16GB ECC RAM
https://www.tomshardware.com/reviews/evga-nvidia-geforce-gtx_1660_super-sc-ultra https://jriver.com/ ), it will be a while before I can hopefully learn to apply DSP filtering to further improve system response before or after acoustical treatment. However, it doing so involves anything like the rigorous math and physics that pervades your discussions on the subject at those links then that's a bridge I have no business crossing.

OTOH, on a percentage basis, how successful would these solutions work for achieving a minimum phase system? https://auralex.com/ ) https://www.dirac.com/live/, https://www.dirac.com/live/bass-control/ and/or https://www.dirac.com/live/dirac-live-active-room-treatment/ ) It obviously works pretty well for a lot of users.

But at minimum, given Pierre’s speaker design notes and Troy’s completed systems, such as these https://josephcrowe.com/collections/finished-speakers-1 , how likely will their crossovers achieve minimal phase?
 
Last edited:
But Cask05 says choosing between those drivers is no solution. Instead, he recommends trying to access among the earliest CD releases of the first generation analog master of a vintage recording and then applying prescribed “mastering” techniques to compensate for the (however unintentionally) destructive effects of excessively applied compression and eq. https://community.klipsch.com/index...taves-audacity-remastering-to-restore-tracks/
I don't believe that I said the above. I never said anything about applying mastering techniques to "compensate for the destructive effects of excessively applied compression". I also said nothing about analog masters. I did say that to avoid the effects of compression in your recordings, you should seek out the earliest CD releases that have not been compressed. Once a music track is compressed, it's basically worthless to hi-fi enthusiasts. You have to go find the same tracks that aren't compressed somewhere else.

Usually, there needs to be a bit of demastering on those older CD releases to negate the bass attenuation that's almost always present. Once a recording has been compressed, essentially, that dynamic range is gone forever. You have to find a CD (or vinyl that can be ripped to PCM) release that doesn't have compression applied.

That is an earlier thread where I responded to a K-forum member's question about minimum phase. If I were going to answer questions about minimum phase today, I think I'd approach the subject a little differently.

Up until perhaps a few years ago, the concept of minimum phase was something that was taught in electrical filtering classes taught to EEs. Nowadays, with the advent of DSP correction in hi-fi reproduction, that concept has been more widely disseminated.

Chris
 
Back
Top Bottom