Jan 032019
 

I have been looking at slightly newer cameras than my ancient Canon 1DS III. There are two big things that have happened since I last took a serious look at cameras :-

  1. Serious cameras are increasingly going mirror-less; last time I looked, electronic viewfinders were too low in resolution and suffered too much lag to really replace optical view finders. 
  2. So-called “medium format” digital cameras are becoming slightly less expensive.

Funny thing is that whilst the “film vs digital” argument has gotten a bit quieter, it is still bubbling away. And as a solely digital photographer, my position on “film vs digital” is simply: it is the final print that counts however you got there.

Back in the days when film was the only viable choice the quality difference between 35mm film and medium format film was dramatic. And similarly between medium format film and large format film.

In the digital world, the difference is more nuanced, and there is more choice in the size of the sensor (“film”) – Micro 4/3 (which is equivalent to the old 110 film format), APC, “full frame”, and “medium format”.

The least honest phrase is of course “medium format” – medium format film came in a variety of different sizes; all of which are actually larger than the medium format digital sensors. The largest “medium format” sensor is approximately 54mm x 40mm; the smallest film medium format is 60mm x 45mm.

Comparing digital and film sizes is pretty irrelevant; with film, quality is directly proportional to size whereas with digital many factors contribute to quality; sensor size being just one.

Part of that quality increase in size is simply down to the increased cost – if you have to make a digital sensor expensive because it is big (fewer sensors per wafer and a higher proportion of them won’t meet the quality standard), then you need to make it better in quality or nobody will buy it. Of course there are also scientific reasons why a bigger sensor is better – or the fancy car priced Phase One cameras wouldn’t have big sensors.

But back to digital sensor size – let us stop calling so-called medium format sensors “medium format” and come up with a new phrase – perhaps “super-frame” and give the crop-factor – 1.67 or whatever it is.

The Windsurfer
Jun 052014
 

There are many different ways to listen to music, and many different ways to listen. The old way to record music is on huge discs containing an analogue recording of the music on both sides – what is now called the Long-Playing record (or LP). It’s called long-playing because it is an evolution of other disc-based formats that were capable of holding less music.

The not quite so old method of holding music is the Compact-Disc (or CD) which uses stores the music digitally rather than in analogue form.

The newest method is the “digital download” and is itself as complex as all the previous methods put together. Because the playback of digital music recordings is basically software, it is possible to introduce new “formats” every so often – so we have MP3, Ogg, FLAC, etc. Most of these formats compromise on music quality in favour of smaller file size so they can be downloaded quicker and you can fit more of them on the average portable music player.

However for the purposes of comparing with LPs and CDs, I will only consider high-resolution FLAC files which do not compromise on quality over file size. Specifically FLAC files with a better than CD quality, and yes they do exist.

When considering music playback, it’s worth remembering that good music always sounds better than bad music; an improved playback merely makes good music sound better. And there are diminishing returns.

Having said that, which of the three is better: LPs, CDs, or digital downloads? The answer is complex, and depends upon the what you want to get out of an album. It also depends on the manufacturing quality – a poorly pressed CD on inferior materials does not compare well with a good CD; and perhaps even more so for an LP.

The Physical Experience and Longevity

In terms of an object to be appreciated in addition to the music nothing beats the LP. Excluding the disc itself, the album art is large enough to be fully appreciated. This is for many an important consideration, but personally I buy recorded music for the music and not for the object d’art that it comes on.

In terms of longevity, nothing lasts forever, but a DRM-free digital download is perhaps somewhat surprisingly in the lead here. Providing that you have a good backup regime, there is no reason why a digital download cannot be passed on to your off-spring.

On the other hand, LPs are subject to various forms of damage that can occur by chance every time you play one – even the playback causes a very tiny amount of damage unless you use a laser-based record player. And every bit of damage (or dust) causes a deterioration in the quality of the playback.

CDs have certain advantages in this respect because it is digital and there is built-in error correction so smaller errors get corrected.

However neither physical medium will last forever, and a poorly treated LP is likely to deteriorate significantly in a relatively short amount of time.

Analog vs Digital

very strongly suspect that the preference that some people have for analog has nothing to do with the fact that it is analog. If you play back a quality LP through a high-end system, it may seem to sound slightly better than a CD. But I have listened to very high quality digital rips of LPs played back on a high quality digital music player, and have heard the same effect – even including the added “warmth” that LP playback supposedly gives you.

analog-digital

The graph above is a very exaggerated demonstration of the difference between a “perfect” analog signal (which frankly doesn’t exist) and a very rough and ready digital signal. The distance between the green lines is representative of the sample rate – how many times per second a digital value is measured. The vertical scale of the green lines is representative of the number of bits used to measure each sample – in this case it is far too few leading to a dramatic staircase effect.

The digital version can be improved by increasing the sample rate and increasing the number of bits for each sample. Increase it enough, and it is impossible to distinguish between the digital and analog versions.

But increase the sample rate to 192KHz and the resolution to 32-bits, and you are talking about some really rather large files. Which is why CDs are just 44KHz and 16-bits; at the time CDs were developed, it was felt to be a reasonable compromise between quality and the need to keep storage requirements down.

One thing that is often overlooked is that CDs contain a lot of error correcting data – 20% of the storage capacity is error correction data. This allows the CD player to correct any read errors that occur (up to a certain level). Nothing else has this level of error correction with the exception of a digital download if it is held on an error-correcting file system (such as ZFS).

This section implies that analog is “perfect”. In case anyone believes that, try reading up on wow, and flutter. And indeed the full details of analog vs. digital. Whilst analog in general is no worse than digital, the specific analog method used for LPs is inherently less detailed than the amount of detail available in CD quality audio and indeed high-resolution digital downloads.

The interesting thing about the so-called “warmth” that LPs provide is that this is effectively making a virtue of the weakness of LPs. The old terms for LP warmth are “muddy bass” and “rolled off high end”. Listening to the warmth of LPs may well be a better experience but that does not make LPs better for music playback.

Digital Compression

Judging by some of the snake oil around digital music, compression is not very well understood – for example. there is a fairly popular myth that the audio from FLAC can be improved by uncompressing into WAV format and then playing it back. There is possibly a very tiny advantage in doing that if you happen to have very poor quality digital equipment, but the advantage would be miniscule. That advantage would be simply down to the hardware/software being unable to uncompress and play back the audio stream at the correct speed all of the time.

There are essentially two different forms of digital compression – lossless and lossy. The former can be uncompressed to produce a digital stream exactly the same as the original, and the latter does not. Lossy compression is much smaller in size, so it is better suited to small portable music players, although there is no reason why small portable music players have to have such small capacities for storage (my current portable player has 256Gbytes of storage and my previous portable player had 240Gbytes of storage).

Because lossy compression loses some of the original data, it approximates the original data – ideally in a way that is indistinguishable from the original. However it is possible that extreme compression – as found in 128kbps MP3 files – will introduce artifacts that sound “wrong”.

Summary

When you come down to it, music is not about digital or analog, it is about the music. And to many quality is also a desirable aspect. This can be obtained with analog or digital, but to a great extent digital is more convenient.

If you are after quality, there is no need to switch to analog to get it – digital music can be just as filled with quality as analog providing that you do it properly.

But most importantly, enjoy the music!

Jul 152009
 

Every so often I encounter a discussion on whether film is better than digital or digital is better than film, which usually degenerates into someone mentioning large format film and someone else mentioning the convenience of digital (or even the convenience of film). It’s all balderdash (and I wrote this post just to use that word … not!). More or less.

When making images (which is what photography is all about after all) it does not matter whether you use film or digital, because using either you can just occasionally produce jaw droppingly good images. Indeed for many such images, the quality of the source does not matter too much as you will be concentrating on the subject rather than the relatively minor “issues” with the image quality such as film grain, ISO noise, chromatic aberration, etc.

What does matter is using whatever makes you comfortable. I cannot shoot film because the thought of actually paying money per shot makes me freeze up. Exposure bracketing ? Forget it. Others cannot shoot digital because computers fill them with horror (and I can certainly understand that!!).

For me, digital is better. For those others film is better.

What counts is the final production – the image, and not the mechanics of how it came about.