Let's somebody in here please explain to me how VBR is bad, except for the fact that it is not well received by the Pio players.

Mathematically, high quality variable uses the 320 bitrate whenever a dynamic portion requires it, but it goes to a lower rate in the rest of the song that doesn't require the full 320 quality. If the mp3 is coded properly, I personally can't tell the difference from any 320k to a high quality variable.

I have no problem adjusting to the pioneer players, as I still prefer cds. But please, all this talk of "crappy quality vbr"??? I am wondering what all this mischief is about.

I don't know the technical side of it since I'm also learning I just know that almost every CD Turntable has less chances of read error using CBR vs VBR

so I'm also listening and I have read few Google links if I get a chance I'll drop those links here if I'm allowed. Thanks

Ever set an audio compressor to act too high too fast with a very long fall-off? That's what a bad VBR encoder will do - overcompress portions of the audio where it shouldn't.

In theory, VBR makes great sense - IN THEORY. What it's SUPPOSED to do is analyze the audio as it's compressing it and apply an appropriate amount of compression to it based on the content. A silent portion of audio could have a very high compression (say 128k) whereas a very full passage would be low compression (say 320k). That's an extreme example for bitrates, most create a smaller bracket much closer to a "center" bitrate, partially reducing effectiveness of the concept.

Why is it bad? It takes a lot of horsepower for the encoder AND decoder to jump around all over the place in order to write / read a VBR file.

As I said at the top, many VBR encoders are sh!t and they act too late - they wind up overcompressing things that shouldn't be and undercompressing lesser passages.

Bottom line - the tiny amount of disk space you will save isn't worth the loss of audio quality and sacrifice of certain functions with DJ players (and some software).

Well written, pulse, I am just curious as to what was behind the poor quality vibe. Yes I agree, if encoded with piss poor software, you wind up with cymbal hits "behind the curve".

Even with optimized settings and a very good VBR encoder, I still don't feel it's worth the extra horsepower required for both encoding and decoding.

Stick with CBRs, storage media is cheap.

Yeah, now it is.

What exactly is this horsepower you speak of? Could you explain to me how the MP3 format works such that VBR becomes an issue?

In my experience, decoding 320kbps CBR is more costly in terms of resources than MP3. MP3 blocks are MP3 frames and the decoder has to do very little to switch from decoding a 32kbps frame to a 320kbps frame. However, it does take time and energy to read a 320kbps frame. As the 320kbps frame is larger than the others, it requires more resources to decode.

EDIT: Wall of verbiage removed.

The point of MP3 ( I think) is to reduce filesize for storage and transmission purposes. There are many problems with encoders out there who do a bad job encoding MP3's. I have my opinions on Pioneer's VBR support, but suffice it to say that REGARDLESS of whether it does or does not use more horsepower to decode a VBR than a CBR, in my opinion you're much better off with a CBR than a VBR because you're less likely to have problems with it when using Pio gear.

-r-

This message has been edited. Last edited by: RyanJ,

<EDIT> Flamebaiting on your first post nets a banning. </EDIT>

This message has been edited. Last edited by: Pulse,

quote:

Originally posted by Pulse:
Even with optimized settings and a very good VBR encoder, I still don't feel it's worth the extra horsepower required for both encoding and decoding.

Stick with CBRs, storage media is cheap.

I am going to ask the same question that Canar asked: What is this horsepower. Is your computer powered by a gasoline engine or something? The truth is that VBR implementation has been optimized for many years now in terms of encoding and decoding. My 2GB 2.5G iPod shuffle is rated for 12 hours of audio playback on a single charge. I continually get about 13.5 hours when playing VBR mp3 files (~170kbps VBR) and VBR AAC files (128-160kbps VBR). I have tested 320kbps CBR files against -V 0 (~280kbps VBR) Lame mp3 files on my lowly iPod shuffle. I was able to get 12.5 hours when using 320kbps CBR and 13 when using -V 0. So again, what is this horsepower? My iPod shuffle used more battery power (and hence, more processor power) to decode a CBR file than a VBR one of the same perceived quality.

Additionally, a constantly powered device such as the Pioneer line of DJ products should be able to handle VBR files just fine. A 320kbps mp3 file requires more processing power to decode. So, if they can handle 320kbps CBR files then they can definitely handle higher (and) bitrate VBR files. My iPod shuffle pretty much runs on a rechargeable watch battery and it can play VBR files for 13-13.5 hours.

If storage was truly not an issue then everyone would use lossless/wav and be done. My point is that it is nonsense to think that VBR mp3 files are worse off than CBR mp3 files. Some devices (such as Pioneer DJ equipment) cannot play VBR mp3 files. That means that they are not mp3 compliant and not true mp3 compatible devices. So one cannot encode their entire lossy library trying to worry about these non-compliant devices. I have a 1G 16GB iPod touch, 5G 60GB iPod, 2.5G 2GB iPod shuffle, 30GB Halo 3 Zune, Xbox 360, PS3, Nintendo Wii, two PocketPCs, Sony digital camera, Honda Civic with mp3/wma six speaker sound system, Toshiba DVD player, and a el-cheapo no name 1GB mp3 player that I use as a USB thumb drive. They all play VBR mp3 files just fine. There is absolutely no excuse for a professional company such as Pioneer to not support part of the mp3 standard (ie VBR).

Go ahead and ban me if you feel that this post is flamebaiting. I see plenty of "flamebaiting" in this thread already with utter nonsense being stated as common truth. That should be enough to ban some members but I guess not. People should head on over to the official Lame wiki page if they want to understand the truth behind VBR encoding and decoding.

Battery life is not an accurate litmus test of required power for decoding.

A VBR decoder works more to vary the bitrate of decoding than a CBR decoder does, which chugs along at the same bitrate for the entire file.

The reason Pioneer won't play the the VBR files is there's more chance for variation in the decoding process resulting in inaccurate playback pitch and time.

Your previous post WAS flamebaiting and if I see you leaning that way again, I'll do the same.

If battery life isn't a good indicator then please tell me what is. Would you consider the following test to be proof enough?

I've carried out tests whilst playing back CBR128 and ~175Kbps VBR MP3 files on a fairly typical PC and, according to Windows Task Manager, the CPU occupancy and RAM used for both Winamp and Foobar2000 remained unchanged regardless of which file I was playing. This infers to me that the decoder isn't having to try any harder to decode VBR than CBR from where I'm looking.

Are you able to provide any evidence of inaccurate playback pitch and time with relation to a properly encoded MP3 VBR file?

I've transferred encodings of the same song in CBR128 and LAME VBR -V3 to a cheap Chinese MP4 player and recorded them back into my PC via the soundcard line-in using a free software package called Audacity. The playback times were identical to the nearest millisecond and expanded plots of both waveforms at various points in the song showed no measurable variance in pitch either.

The VBR file seeks correctly and always shows the correct time in relation to position when playback resumes on this cheap Chinese MP4 player. The same applies to Winamp and Foobar2000 too.

I'm not setting out to disprove what you're saying, I'm just politely asking you to provide evidence of your claims, because I don't think they hold water when a properly encoded MP3 VBR file is played back on competent equipment.

As was said in the post that was deleted earlier, VBR has been a part of the MP3 standard for a very long time, at least since 1993 as far as I can ascertain, so why does some Pioneer hardware still struggle with it 15 years later when it's been a part of the ISO/IEC international standard IS 11172 for all of that time?

In relation to your claim earlier regarding sluggishness in the response time for bitrate selection in VBR, the LAME encoder doesn't behave in that way. It monitors its own output constantly for level of quality and selects a bitrate accordingly. There is also no need to deliberately restrict the minimum bitrate to 128Kbps either. Digital silence is encoded at 32Kbps with LAME in VBR mode and all standard MP3 bitrates are included in the range of acceptability if the encoder deems them sufficient.

I respectfully suggest that you read the LAME WIKI to become more conversant with the facts regarding LAME. It's the most competent MP3 VBR encoder beyond a shadow of a doubt and is widely regarded by those in the know as providing perceptually transparent results down to an average bitrate of around 175Kbps, possibly lower.

This message has been edited. Last edited by: StormChaser,

For starters, how is it you are unable to post under the same account each time?

I can understand why Pioneer decided not to support anything below 128k -- it sounds like crap. If it were my choice, I'd not support anything below 256k.

A CDJ is not some "cheap Chinese MP4 player". It is a precision player and much like an F1 car requires specific fuel, bad MP3s will affect the playback of the CDJ. There's more to just the decoding - it takes a lot of processing power for the ability to decode ahead of the playback at variable speeds and provide accurate buffer data flow regardless of what the user does with the platter, pitch or hotcues. Variable bitrate files require variable bitrate decoding.

While you or others may claim "perceptually transparent results" with a 175k VBR MP3, I can provide equal results for the general public not being able to tell the difference between a 4MP 6MP or 12MP digital image printed at 8 x 10. But show it to a professional who knows what to look for and he can peg each of them, just as a DJ should know what to listen for in order to hear the differences between a 128k CBR and a 320k CBR.

I'm not quite sure what the whole point of your tirade is here besides the question of "why doesn't Pioneer support VBR", to which I answer: Go ask the engineers in Japan, I'm here to support the users.

For starters, I resent you accusation of me being some kind of multiple spammer. I AM an individual, and this is only my second post here. As you have the power to edit others posts, I can only assume that you are either an admin or a mod on this forum. Try behaving like one rather than randomly insulting people or Pioneer will be receiving an email regarding your ungentlemanly conduct.

Concerning your other points, prove that an MP3 VBR encoding is "bad". By definition, any VBR file that complies with internationally agreed standards is "good". If it's also perfectly capable of providing perceptually transparent results in open, public group listening tests, then this reinforces the fact that it is a good MP3 file.

Prove that it requires more processing power to decode a VBR file than a CBR file. I find no evidence anywhere to support this claim either via extensive Net searches or personal experimentation.

Also, prove that the use of bitrates as low as 32Kbps when appropriately used during the VBR encoding process damages the sound quality. 32Kbps will only ever be used by the LAME encoder during periods of digital silence anyway, so it makes no difference. Just saying things without backing them up with factual evidence means nothing to anybody. There is plenty of evidence on the Net to prove that it does no harm whatsoever.

Saying that a specific MP3 player "requires a specific fuel" is complete nonsense. An ISO/IEC compatible MP3 file is a correct and proper MP3 file and will play back perfectly at all times on any ISO/IEC standards compliant MP3 player. You inferring that the file is broken in some way doesn't alter the fact that, according to the people who created the MP3 standard in the first place, it isn't.

I did read your response with regard to my doubts regarding pitch variance and time errors resulting from the use of VBR MP3 files, and I do see other evidence of that, but it only seems to happen on non-ISO/IEC compliant MP3 players, unless you know otherwise. If any postprocessing in the form of scratching, hotcueing, etc takes place, surely this happens after the MP3 decoding has taken place so should have no bearing whatsoever on anything relating to the original decoded audio stream?

The resultant audio from the MP3 decode will presumably be buffered as raw audio before it ever gets to the stage of being interacted with by the DJ, no? As the final audio buffer size required for a block of VBR data is going to be identical to that required for a block of CBR data assuming that they're both internally decoded to 44.1kHz 16-bit (guessing?), what difference does it make whether the source was VBR or CBR?

As this appears to be a genuine Pioneer forum, if you're not capable of answering the questions posed above with a degree of technical competence, may I respectfully request that you approach someone who can do so on your behalf? Maybe I'm mistaken in presuming that your title of "Pioneer National Trainer // Product Specialist" puts you in a position to answer these technical questions with the professionalism I'd expect.

Making blanket statements such as, "I can understand why Pioneer decided not to support anything below 128k -- it sounds like crap. If it were my choice, I'd not support anything below 256k." proves that, even though you may understand the equipment you deal with, you have no understanding whatsoever of the internal workings of MP3. Please go and learn something for the sake of all other members of this forum and stop spreading what are effectively lies generated by your own ignorance.

I'm very disappointed with the reception I've had here to be honest. Others spout jibberish here and are left alone. Talking actual facts and figures seems to lead to personal attacks and abuse. It's already happened three times in this one thread alone to three new members. Why is that?

PS I have saved a copy of this post to attach to an email to Pioneer should it mysteriously "disappear".

This message has been edited. Last edited by: StormChaser,

I never accused you of being a spammer - obviously you need to learn your terminology. This thread was dormant for 6 months then all of a sudden we have posts from users Canar, BDX, ThisIsNot and StormChaser - ALL registered to the forums on the same date, all from different global IPs, all of which not putting their true locations in their profiles. That's called "trolling".

Given your vehement defence of VBR, I'd have to guess that you have all come from some audiophile forum. One person found the thread and posted it back there and rallied the troops to come and post your minds here.

Just because I'm a product specialist doesn't mean I know of all the technical inner workings of the players. I have the knowledge of all the functions and features of the units - I demonstrate them and train users and dealers on their use and operation.

I have a car and I know certain things about its technical inner workings but that doesn't mean I know what the compression ratios are for my engine. I don't know the gapping size for my spark plugs. There is plenty which I don't know about my car, just as there's plenty about the Pioneer products I don't have intimate knowledge of as I'm not an engineer.

My personal experiences with VBR have never been good. I've been encoding MP3s for quite some time and I can honestly tell you that when I started encoding to VBR, I found it to take longer than CBR encoding and I was never as happy with the results of the sound quality of VBR as with the CBR equivalent. Quite often with VBR I would find that portions of that audio where the bitrate increased were "overcompressed"; the encoder acted much like a low-speed limiter/compressor (audio device, if you aren't familiar with it) and the passage of audio would thus sound a tiny bit muddier, then clear-up. Based on those experiences, I moved to CBR and never looked back.

It's not ignorance, it's experience and preference. I like learning - I would much rather have knowledge of a subject to be able to discuss it than spouting off about something I know nothing about. So while I don't claim to be an MP3 expert and I can certainly stand to learn more, I will tell others about what I have learned works for me and what doesn't.

I am allowed to state opinion - that's my right. You are allowed to argue against my opinion - that's your right. If you are insistent on learning more about the inner technical workings of the Pioneer products, I encourage you to send me an email and I would be more than happy to pass your email along to an engineer at Pioneer who could best answer your questions or discuss these issues at length with you myself. As I said, I am always willing to learn and it sounds like you could teach me something about MP3 just as I could teach you something about Pioneer DJ gear.

Thanks for your reasoned and civilised reply. Now I think we're both getting somewhere, although I still feel as though I'm being unfairly labelled as a troll. I just want people to have the opportunity to learn the truth rather than misinformed half-truths based on historical experience when lossy encoder performance is in a perpetual state of flux, and usually for the better. I don't consider that to be the act of a troll, rather the act of a potential friend to the forum.

I'm glad that you admit your limitations, as do I. It would be unreasonable of anybody to expect you to know everything about all matters pertaining to MP3 encoding as not many do. It takes a particular mindset and a fair amount of time to assimilate all of the information involved, not that this is necessary in order to make full use of an encoder for the purely practical purpose of encoding an MP3 file. Reading and mentally digesting a good article written by experts on the subject is normally sufficient, and I respectfully point you HERE if you have a personal or professional interest in learning the basics of "driving" LAME as officially recommended.

My personal experience of VBR MP3 encoding is largely limited to the LAME encoder which I've used pretty much exclusively for roughly three years now. Many tests have been undertaken by users with demanding ears and equipment and perceptual transparency is widely (although not universally) agreed to be attained with a setting of -V3. This equates to a typical average bitrate of around 175Kbps over many tracks but still has access to the entire range of standard bitrates from 32Kbps to 320Kbps as the encoder deems necessary.

According to the WIKI entry, and I quote, "Unlike other MP3 encoders which do VBR encoding based on predictions of output quality, LAME's default VBR method tests the actual output quality to ensure the desired quality level is always achieved.". I'm unaware of any tendency to undershoot or overshoot except at the extreme ends of the -V scale which ranges from -V9 (lowest quality) to -V0 (highest quality). -V5 through -V2 seem to cover most people's needs.

To the best of my knowledge, all files produced by LAME in VBR mode using the standard -Vx switch are ISO/IEC compliant. I'd be very interested from a technical point of view to know how well the Pioneer DJ decks handle a file produced in this way and would very much appreciate it if you could find the time to test this. If you have a PC laying around running Windows then Foobar2000 is an excellent application for creating VBR MP3 files with LAME from an audio CD. Many guides exist on how to do this, although you can probably suss it out yourself quite quickly.

Regarding the handing over of my email address, nothing personal, but I don't give it out to people I don't know. Maybe one of your tech guys could drop a post in this thread (or a new one) linking us to a flow diagram of the process that takes place in terms of digital signal handling and processing when an MP3 CD is dropped into one of your machines, and how it varies in relation to the same task(s) performed from an actual audio CD. I'm itching to know what's going on to be honest, and I don't think a flow diagram would give away any trade secrets.

Don't feel obliged to follow up anything I've said above if you don't have the time or personal interest in doing so, but I'd certainly appreciate any more information coming in my direction. I'm always keen to learn something new.

Pulse, I've been paying attention and discussing with them over on the forum they come from. These boys seem to know what they're talking about.

I know I can't post the link to over there, so I'll just tell you next time I see you online (come online you bugger).

In one of the posts someone does some VBR vs. CBR encode time trials - and shockingly enough, VBR comes out on top. I'd say that encode time IS a fair litmus test of processor juice required; at least for the encode process.

Most of the members over there concede that 320kbps IS better than a properly encoded VBR, but many of them believe that MP3 is so far from perfect that the mild sonic quality difference you get from using a file that is larger (aka the CBR) isn't worth the extra space. (After all, saving space IS the point of MP3). This is all on newer encoders; they ALSO concede that early implementations of VBR encoders hosed the works up pretty bad and gave VBR a bad name.

Also, the forum in question has links to listening tests that are conducted regularly on many different samples. It uses a separate program that doesn't let you know what you're listening to - you post your thoughts and after a set time limit the results are revealed.

I'm going to take one to see if I can tell.

My primary argument against VBR's is pretty much that there are so many that are broken and/or are bad that make things break. Therein lies the problem - if everyone encoded their own MP3's, we might not have this issue - but very few people actually do (*cough*limewire*cough*).

I'm following this whole thing with great interest and I'm learning a lot.

-r-

Well said all. A few, very few, thoughts.

Interesting to hear ryan, who knows his stuff, say that VBRs come out on top?? I believe it, but need to get over my bias (due to crap i've gotten on line in the past- bad sites to begin with). Pulse and I can go over this at NAMM with some of the engineers and see if it's on the horizon.

Stormchaser, my son's favorite Weather Channel show, are you a DJ or a sound guy or gal? yes, when a thread sits quiet and then 4 new users come on with BIG brains and statements, it's a red flag for the 2 moderators (Pulse and I). But it sounds like all is well and no need for me to go on about anything else.

Please continue to play nicely all......

Jay, note that I said that VBR comes out on top for processing, and that with the exception of 320kbps CBR (which most people seem to agree that they can't tell the difference between it and a properly encoded VBR), VBR sounds better at a lower filesize.

Note, however, that the key is "properly encoded". Anyone can make any MP3 sound like crap if they try hard enough (or, in some cases, don't) - even a 320kbps CBR.

-r-