Your site says that you "losslessly compress images, video, and audio" - I poked around in the code a little and didn't find the compression stuff, but it doesn't make sense to me that you can do that. There is some optimization that you can do on PNGs, but for most media encoded with a lossy method you shouldn't be able to achieve a smaller filesize without reencoding with another lossy method (and by definition losing more information, althoughh it might not significantly decrease the percieved quality)
Actually, we just took some of that out to reduce processing times for users. We're going to overhaul the backend processing system so that we can process some things asyncronously, and then we'll put all that code back.
However, we do losslessly compress some things. We run PNG files through optipng and JPGs through jhead to strip out EXIF data, but most interestingly, we run GIF files through ffmpeg and serve them up with HTML5 video [1]. We usually get between 500 and 2000% faster for GIFs.
However, I agree that it's a little misleading, since we don't do it for every kind of media, and an ideally compressed file cannot be compressed further. I've been considering rewording it.
As someone who does media compression pretty much daily, your marketing spiel really came out to me badly for the same reasons. While you can optimize JPGs and PNGs in a lossless fashion, you really can't do the same for audio and video. You might be able to make them smaller without losing (much) perceived quality (aka do transparent compression), but you're still doing lossy compression. Same with converting gifs to videos - while the biggest loss certainly happens in the making of the original gif and while you can get a humongous increase in compression quality, converting it to VP8 is still lossy.
And speaking of which, saying that you can get "1000-3000% faster for some files" is also pretty dishonest. You can pretty much claim those kind of numbers with only one of the many formats you support (and one where you're not doing lossless compression), but the way you word it makes it sound like you could get it for potentially anything.
All in all, I'd really suggest you rewrite the description to be more honest.
Two other things I noticed: I can't seem to select text on the homepage (on latest Chrome). Your icon also looks rather similar to that of Miro: http://www.getmiro.com/
I agree. We originally started to simply deal with GIF compression (MediaCrush was previously known as gifquick) and we have less ground to stand on with respect to spectacular compression with support for more formats. I'll reprioritize the "rewrite the spiel" task thanks to feedback from HN.
As for selecting text on the home page, not much we can do about it. We force your focus into a contenteditable div to allow you to paste images/URLs directly into the page.
Regarding the icon, it just so happens that I had the sneaking suspicion that I had seen it before when the designer presented it to me. So I ran a reverse image search and looked at about a thousand similar icons. I didn't find Miro - if I had, I probably would have asked for it to be different. In any case, I think it's different enough that I'm not worried.
We all know how H.264 is great. You don't need to explain it again.
And you should fix it right now. I already got super-bad first impression. Currently your product is just marketing junk shit to me. And you have no way to fix my impression because I won't review your product ever again.
> We all know how H.264 is great. You don't need to explain it again.
You know that, but our intended audience isn't the HN crowd.
I'm sorry to hear that you got a bad impression of the site, but even without any fancy speed tricks (and there are fancy speed tricks), it's still an open-source, privacy-centric media hosting site that I think is pretty damn great.
It could be great. And that exaggerated text voids everything. That's why I told you to fix it ASAP. Before that text make you to lose any more people.
