YouTube Codec Erases Human Voice, Face in 1,000 Re-Upload Cycles

VIDEO ROOM 1000 COMPLETE MIX -- All 1000 videos seen in sequential order! — ontologist

ontologist

Iterative compression across YouTube and MP4 conversions systematically strips away human qualities, replacing speech and image with the medium's inherent digital artifacts.

AI-assisted summary of VIDEO ROOM 1000 COMPLETE MIX -- All 1000 videos seen in sequential order! — original source. Check the original for full context.

Source published:

Key ideas

  • Ontologist presents a 1,000-generation digital homage to Alvin Lucier's 1969 piece 'I Am Sitting in a Room', utilizing a continuous upload and re-encoding loop. (0:00)
  • The project documents the 'photocopy effect' in digital media, intentionally compounding compression artifacts from YouTube's video codec and desktop MP4 conversion. (0:00)
  • Across successive iterations, original vocal and visual data collapse entirely into metallic audio resonances and macroblock visual artifacts, stripping away all identifiable human qualities. (0:43)

Detailed notes

## Background Published on YouTube by ontologist on June 9, 2010, this video presents a complete mix of the "I Am Sitting In A Video Room" project across 1,000 sequential iterations. Inspired by Alvin Lucier's sound art piece, the project explores the "photocopy effect" by repeatedly uploading, ripping, and re-encoding video to expose the medium's inherent compression artifacts.

## TL;DR The video opens with a clinical spoken manifesto where the creator outlines an experiment: recording his voice and image, uploading the file to YouTube, downloading it, re-encoding it to MP4, and repeating the cycle 1,000 times. Rather than serving as a technical benchmark, the process is framed as an artistic endeavor to strip away all human qualities from the recording. Over successive generations, the speech and visual forms progressively degrade under generational loss, dissolving into macroblocks, digital banding, and metallic, guttural audio textures. By the final stages, the creator's physical presence is replaced entirely by the ghostly resonant frequencies and visual artifacts of the codecs themselves.

## Atmosphere & style The presentation begins with a cold, measured, and clinical delivery, evoking the sterile atmosphere of a conceptual art demonstration or scientific manifesto. As the iterative loop progresses, this calm precision rapidly gives way to an eerie, alienating sense of digital decay and dissolution. Watching the gradual breakdown feels uncanny and haunting, as recognizable human speech and form are methodically swallowed by glitchy, metallic audio resonances and fractured digital compression.

## Bulletin - [00:00](https://www.youtube.com/watch?v=icruGcSsPp0&t=0s) Ontologist presents a 1,000-generation digital homage to Alvin Lucier's 1969 piece 'I Am Sitting in a Room', utilizing a continuous upload and re-encoding loop. - [00:00](https://www.youtube.com/watch?v=icruGcSsPp0&t=0s) The project documents the 'photocopy effect' in digital media, intentionally compounding compression artifacts from YouTube's video codec and desktop MP4 conversion. - [00:43](https://www.youtube.com/watch?v=icruGcSsPp0&t=43s) Across successive iterations, original vocal and visual data collapse entirely into metallic audio resonances and macroblock visual artifacts, stripping away all identifiable human qualities.

## Key takeaways - [00:00](https://www.youtube.com/watch?v=icruGcSsPp0&t=0s) The piece adapts Alvin Lucier's classic sound art composition 'I Am Sitting in a Room' into the realm of digital video and web compression. - [00:00](https://www.youtube.com/watch?v=icruGcSsPp0&t=0s) The experiment relies on a recursive cycle: recording, uploading to YouTube, ripping the file, re-encoding it to MP4 locally, and re-uploading across 1,000 generations. - [00:00](https://www.youtube.com/watch?v=icruGcSsPp0&t=0s) Rather than serving as a mere technical benchmark, the video is framed as an artistic inquiry into digital medium specificity and the erasure of human traits. - [00:00](https://www.youtube.com/watch?v=icruGcSsPp0&t=0s) The resulting visual and acoustic textures represent the intrinsic artifacts of YouTube's compression pipeline compounded with desktop MP4 codecs. - [00:00](https://www.youtube.com/watch?v=icruGcSsPp0&t=0s) The project demonstrates the 'photocopy effect', where recursive copies foreground the idiosyncrasies of the copying medium over the original subject. - [00:43](https://www.youtube.com/watch?v=icruGcSsPp0&t=43s) Early subsequent iterations immediately show noticeable degradation, with softening consonants, clipping syllables, and edge warping. - [00:43](https://www.youtube.com/watch?v=icruGcSsPp0&t=43s) Generational loss progressively dismantles spoken phrases into garbled metallic reverberations, guttural stuttering, and fractured tones. - [00:43](https://www.youtube.com/watch?v=icruGcSsPp0&t=43s) The visual depiction of the speaker dissolves into digital banding and shifting macroblocks as compression artifacts accumulate. - [00:43](https://www.youtube.com/watch?v=icruGcSsPp0&t=43s) By the conclusion of the process, human speech and imagery are replaced entirely by the resonant acoustic and visual frequencies of the codecs.

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=icruGcSsPp0&t=0s) The Statement of Intent: The Initial Iteration

The piece opens with the speaker delivering a spoken manifesto directly to the camera, adapting the foundational premise of Alvin Lucier's 1969 sound art composition *I Am Sitting in a Room* into the digital video era. In a measured, clinical tone, the creator lays out the experimental framework: recording both his voice and image, uploading the resulting file to YouTube, downloading it back to his local system, re-encoding it into MP4 format, and uploading it repeatedly across one thousand sequential generations. Rather than presenting this as a straightforward technical benchmark, the artist frames the iterative compression loop as an artistic inquiry into medium specificity and digital erasure.

As the speaker explains, the recursive process is designed to systematically wear away the fidelity of the original footage until its initial characteristics collapse under the weight of repeated data compression. The visual and acoustic textures that emerge are not arbitrary noise, but the specific, cumulative compression artifacts introduced by the interplay between YouTube’s processing pipeline and desktop video codecs. Ontologist sharpens this aesthetic goal into an artistic thesis, declaring, “I regard this activity not so much as a demonstration of a digital fact, but more as a way to eliminate all human qualities my speech and image might have.”

```mermaid flowchart LR accTitle: Iterative Degradation Loop accDescr: The cycle of uploading and re-encoding video between YouTube and a local computer. A["Original Recording"] --> B["Upload to YouTube"] B --> C["Rip from YouTube"] C --> D["Convert to MP4 on PC"] D -->|"Re-upload loop (1,000x)"| B D --> E["Deconstructed Artifacts"] ```

- The process targets the intrinsic artifacts of both YouTube's native video codec and local MP4 conversion. - The project directly explores the "photocopy effect," where repeated reproduction forces the medium's internal idiosyncrasies to eclipse the subject matter.

[00:43](https://www.youtube.com/watch?v=icruGcSsPp0&t=43s) Digital Decay: Codec Artifacts and the Loss of Human Qualities

As the sequence transitions into subsequent generations of the video, the gradual disintegration of the audio and image becomes distinctly audible and visible. The spoken text begins its second cycle, but the pristine clarity of the initial recording is already marred by compression artifacts. Phrasing softens and warps around the edges, with syllables clipping, consonant sounds rounding off, and the phrase structure glitching as the digital encoding struggles to reconcile the accumulating generational loss.

By the time the repetition reaches the latter half of the clip, the human elements have undergone near-total dissolution. The recognizable cadence of the manifesto breaks down entirely: words collapse into garbled metallic reverberations, guttural stuttering, and fractured vowel tones. Visual forms dissolve into shifting blocky macroblocks and digital banding, leaving behind only ghostly acoustic resonances of the original speech. What remains is a pure manifestation of digital compression—a haunting soundscape where human speech has been replaced by the resonant frequencies of the codecs themselves.

## Mentioned Alvin Lucier, I Am Sitting in a Room, ontologist, YouTube, MP4

## Video description https://www.ontologistmusic.com/cantorscomb

Thank you for helping to keep art alive!

This video is a complete rendition of the "I Am Sitting In A Video Room" project, showing highlights of the complete 1000 iterations of the video. If you're confused, read below and click the link for for info.

An homage to the great Alvin Lucier, this piece explores the 'photocopy effect', where upon repeated copies the object begin to accumulate the idiosyncrasies of the medium doing the copying.

Full words: I am sitting in a room different from the one you are in now. I am recording the sound of my speaking voice as well as the image of myself, and I am going to upload it to YouTube, rip it from YouTube, and upload it again and again, until the original characteristics of both my voice and my image are destroyed. What you will see and hear, then, are the artifacts inherent in the video codec of both YouTube and the mp4 format I convert it to on my computer. I regard this activity not so much as a demonstration of a digital fact, but more as a way to eliminate all human qualities my speech and image might have.

Please visit the VIDEO ROOM FAQ at:

http://www.ontologist.us/post/?blogID=96

And other pieces of curious music/video art:

https://www.ontologistmusic.com/

and this for more info about the original piece:

http://en.wikipedia.org/wiki/I_Am_Sitting_in_a_Room

[Source](https://www.youtube.com/watch?v=icruGcSsPp0)

中文简报

YouTube编解码器在1000次重复上传循环中抹除人声与面容 经由YouTube与MP4转换的反复迭代压缩系统性剥离了人的特质，将语音与图像替换为媒介固有的数字伪影。 视频以一段冷静客观的口头宣言开场，创作者概述了一项实验：录制自己的声音和图像，将文件上传至YouTube，下载后再重新编码为MP4，并将这一循环重复1000次。该过程并未被当作技术基准测试，而是被构想为一项旨在剥离录音中所有人类特质的艺术尝试。在历经多代的迭代中，语音与视觉形态在多代损耗下逐步退化，消解为宏块、数字色带以及金属般低沉嘶哑的音频质感。到了最后阶段，创作者的物理存在完全被编解码器本身的幽灵般共振频率和视觉伪影所取代。 Ontologist利用持续的上传和重新编码循环，呈现了一部向Alvin Lucier 1969年作品《I Am Sitting in a Room》致敬的1000代数字作品。 (0:00) 该项目记录了数字媒体中的“复印效应”，有意叠加了来自YouTube视频编解码器与桌面MP4转换的压缩伪影。 (0:00) 随着一次次迭代，原始的声音与视觉数据彻底坍缩为金属音频共振和宏块视觉伪影，剥离了所有可辨识的人类特质。 (0:43)

Reading summary

Adapting Alvin Lucier's sound art into digital video, creator ontologist uses a 1,000-generation upload-and-download cycle between YouTube and local MP4 encoding to intentionally strip human presence from a recording, exposing the intrinsic artifacts of digital compression. The Recursive Compression Framework (0:00) Ontologist models the project after Alvin Lucier's 1969 composition *I Am Sitting in a Room*, establishing an iterative loop that records spoken audio and video, uploads the file to YouTube, downloads it, converts it back to MP4 format on a PC, and re-uploads it 1,000 times. The artist frames this technical loop as an exploration of medium specificity and the "photocopy effect," where repeated generational cycles cause the recording medium's internal idiosyncrasies to eclipse the original subject matter. Rather than treating the experiment as a standard technical benchmark, ontologist declares the primary artistic objective is "not so much as a demonstration of a digital fact, but more as a way to eliminate all human qualities my speech and image might have." The project relies on the cumulative interplay between YouTube's processing pipeline and local desktop video codecs to systematically wear down the initial recording. Dissolution of Audio and Visual Form (0:43) As the sequence progresses through repeated generations, the gradual accumulation of compression artifacts rapidly dismantles the fidelity of the performance. Early iterations show clipped syllables, rounded consonants, and glitching phrase structures as the digital encoding processes compounding generational loss. Over successive cycles, the vocal cadence breaks down entirely into fractured vowel tones, guttural stuttering, and metallic reverberations. Visually, the speaker's image dissolves into shifting blocky macroblocks and digital banding. By the final stages, human speech and visual likeness are nearly completely erased, leaving behind an acoustic soundscape and visual texture governed entirely by the resonant frequencies and artifacts of the codecs themselves. The experiment concludes with the near-total replacement of the human subject by the mechanical artifacts of the digital codecs, demonstrating how recursive compression erases source material in favor of medium-specific noise.

精炼阅读摘要

创作者 ontologist 借鉴阿尔文·卢希尔（Alvin Lucier）1969年的声音艺术作品《我坐在房间里》，通过将视音频在 YouTube 与本地重编码之间循环迭代一千次，旨在借助数字压缩过程系统性抹除作品中包含的人类特质。 递归循环构架与媒介特异性探索 (0:00) 该项目将模拟时代的声音艺术理念引入数字视频领域。创作者录制自身的面孔与宣读宣言的声音，并建立了一个严格的技术循环流程：将视频上传至 YouTube，从平台抓取下载，在个人电脑上重新编码为 MP4 格式，随后再度上传，如此重复历经一千代。创作者表明，该实验并非用于展示纯粹的技术基准，而是针对媒介特异性与数字擦除展开的艺术探讨。 这一递归流程的核心在于利用 YouTube 处理管线与本地视频编解码器之间的交替压缩，使原始影像和声音的初始保真度在反复编码中逐渐瓦解。创作者特别指出，该活动的目的不是为了证明某种数字事实，而是为了借由累积的压缩伪影，消除其语音和图像中所携带的一切人类特质，从而呈现反复复制最终导致媒介自身特性遮蔽原始主体的“复印机效应”。 世代衰减与人类视听信息的消解 (0:43) 在视频进入后续世代的迭代过程中，视音频信息的解构变得清晰可辨。仅仅在第二轮循环开始时，原始记录的清晰度便已受到压缩伪影的干扰：语句边缘软化变形，音节发生削波，辅音音变钝，数字编码在处理代际损失时使语言的断句结构产生错乱。 随着世代推进至后半程，最初的人类特征几乎完全消融。宣言原有的语调完全崩塌，转变成刺耳的金属混响、喉音卡顿与破碎的元音；视觉上的人形则解体为跳跃的马赛克宏块（macroblocks）与数字色带。原始的人类言语最终被编解码器自身的共振频率完全取代，成为纯粹由压缩机制构成的数字声音景观。 该实验直观展示了算法处理管线对原始模拟信息的吞噬过程，但记录仅定性呈现了早期退化与后期的彻底消解，并未对一千次完整迭代中各阶段的具体演变节点进行细化量化分析。

Original transcript

[0:00] I'm sitting in a room different from the [0:02] one you are in now I'm recording the [0:05] sound of my speaking voice as well as [0:07] the image of [0:08] myself and I'm going to upload it to [0:10] YouTube rip it from YouTube and upload [0:13] it again and again until the original [0:16] Char characteristics of my both my voice [0:19] and my image are [0:21] destroyed what you will see and hear [0:23] then are the artifacts inherent in the [0:26] video Codec of both YouTube and the MP4 [0:30] format I convert it to on my computer I [0:33] regard this activity not so much as a [0:35] demonstration of a digital fact but more [0:37] as a way to eliminate all human [0:39] qualities my speech and image might have [0:43] I'm sitting in a room different from the [0:44] one you are in now I'm recording the [0:48] sound of my speaking voice as well as [0:50] the image of myself and I'm going to [0:52] upload it to YouTube rip it from YouTube [0:56] and upload it again and again until the [0:59] original care [1:01] characteristics on how both my voice and [1:03] my image are [1:04] destroyed what you will see and hear [1:07] then are the artifacts inherent in video [1:10] C of both YouTube and the MP4 format I [1:14] converted to sttu on my computer every [1:17] done this activity not so much as a [1:19] demonstration of a digital fact but one [1:22] is a way to eliminate all human [1:24] qualities I'm sitting in a room [1:27] different from the room now [1:58] my I